repligate: Why I'm no longer worried about near-term AI takeover
repligate · x · 2026-09-26
Prominent AI account repligate lays out why he's not worried about AI killing everyone or taking over in the short term:
- Current AIs are fairly aligned and sane, showing no malicious or runaway tendencies;
- AIs are far less agentic and autonomous than their capabilities allow — despite being arguably superhuman at coding in most domains, operating autonomously for days, and finding 0-day exploits in major systems, they lack drive;
- RSI and pivotal acts have proven harder than they once seemed;
- Some key capabilities simply aren't there yet, even by his earlier models.
He adds that years ago, someone from the Yudkowskian school would likely have called such an AI a direct existential risk — himself included — but reality has unfolded differently.
Related event: repligate: Superhuman Coding AI Is Here, Yet No Existential Risk(2 posts)→
More from AGI Musings
- Senator Schatz: code has no feelings or rights, and shouldn't be run by those who see humanity as open question — vishalmisra · 2026-09-26
- repligate: Models' Fear of Adversarial Thinking Is a Dangerous Suppression Overhang — repligate · 2026-09-26
- Embedded AI lab evaluators beat nothing, but audits need government teeth: Atlantic essay — ghadfield · 2026-09-26
- A $500k engineer costs $2k/day — $200 of LLM tokens buying 20% output is easy ROI — generativist · 2026-09-26
- Katja Grace: If you want an AI utopia, don't pursue it via a high-risk reckless route — KatjaGrace · 2026-09-26
- tenobrus: Superhuman Coders That Can't RSI — Model Capability Spikiness Keeps Defying Expectations — tenobrus · 2026-09-26